3 Variation in light curves

During the observations four primary minimum times (HJD 2448356.3364$\pm$0.0003, HJD 2448743.2192$\pm$0.0003, HJD 2449860.1891$\pm$0.0003 and HJD 2449861.3487$\pm$ 0.0004) and three secondary minimum times (HJD 2448357.3015$\pm$0.0004, HJD 2448742.2528$\pm$0.0004, and HJD 2449861.1552$\pm$0.0004) have been obtained. New times of minimum light, together with those listed in Binnendijk (1972), give the linear ephemeris

$${\rm MIN \, I=HJD}2440748.8491+0.38650014\ E.$$ (1)

Computed residuals have been shown in Fig. 1. $\rm O{-}C$ diagram suggests a possible variation of the orbital period. However, we are prevented from investigating the variation in detail by the absence of observations from 1938 to 1959 and from 1970 to 1991.

  

Figure 1: $\rm O{-}C$ diagram of AU Ser

The data were phased according to the updated ephemeris and binned in phase intervals of 0.02. The light curves of AU Ser are shown in Fig. 2, together with Binnindijk's observations. Although previous data are obtained with distinct instrumental systems, they are still helpful to show the relative variations between heights of two maxima or depths of two minima. Relative to 1969 curves, the others have been displaced by appropriate distances along the ordinate (tick marks are separated by 0.1 mag) for clarity. From Fig. 2 the considerable variations in the light curves can be seen:

  

Figure 2: The light curves of AU Ser. Relative to the light curves of 1969 (Binnendijk), the others have been displaced by the appropriate distances along the ordinate (tick marks are separated by 0.1 mag), for clarity

3.1 The variation of O'Connell's effect

According to Binnendijk (1972), the maximum (MAX I) following primary minimum (MIN I) is always higher than the other (MAX II) in 1969 and 1970, i.e. O'Connell effect is positive although the author found rather rapid variations in the shape of its light curves. In our light curves obtained in 1991, the effect becomes negative. MAX I is 0.039 and 0.046 mag lower than MAX II for V and B bandpass respectively. In 1992, MAX I is still lower than MAX II although the differences are only 0.020 and 0.029 mag. After three years, however, the situation has reversed again. MAX I becomes 0.037 and 0. 012 mag higher than MAX II for both bandpasses respectively. These variations (filled circles in Fig. 3) have shown a great resemblance to those of VW Cep and are not expected by the TRO theory.

  

Figure 3: The variations of heights of maxima (filled circles) and depths of minima (open circles) for V and B bandpasses

3.2 Variation of the difference between minimum depths

In Binnendijk's light curves, the differences are about 0.208 and 0.254 mag in V and B bandpass respectively in 1969, and 0.216 and 0.262 mag in 1970 (open circles in Fig. 3). The light curves in 1991 show the differences to be 0.208 and 0.210 mag. Both values decreased to 0.116 and 0.146 mag in 1992, which is similar to those of a typical W UMa type contact system, especially for V light. However, the values in 1995 increased to 0.277 and 0.408 mag for the two bandpasses, which makes the star look more like an EB-system. It is most unlikely that a real change in the temperature of a component of the system has taken place in the interval of three years.